7/26/2019 1997.12 Zhu et alnm,m

1/18

EXPERIMENTAL STUDIES OF

ELECTRO

SEISMIC

CONVERSION

IN

FLUID SATURATED POROUS

MEDIUM

Zhenya Zhu Matthijs

W. Haartsen

and M. Naft Toksoz

Earth Resources

Laboratory

Department of

Earth tmospheric and Planetary Sciences

Massachusetts

Institute

of

Technology

Cambridge M 02139

BSTR CT

The

coupling between seismic and electromagnetic waves in a fluid saturated porous

medium

is

essentially controlled by the electrokinetic effect. The inverse effect of the

seismoelectric conversion the electroseismic conversion

is

investigated experimentally

in the laboratory.

The

electric field induces movement of

the

ions in a pore fluid relative

to the solid matrix. The interaction between the pore fluid and the solid

matrix

gener-

ates acoustic waves known as electroseismic waves.

Our

studies confirm experimentally

that

the

electroseismic conversion

at

ultrasonic frequencies is the electrokinetic effect in

nature. In measurements with a homogeneous rock cylinder P and S wave transducers

receive respectively the P and

components of the extensional

and

flexural waves

generated by an electric pulse with ring or parallel electrodes when the electrodes are

on the surface or inside a porous medium. The electroseismic waves are measured in

layered models made of sandstone or artificial materials to determine

the

area where

the

electroseismic waves are generated. Further experiments with

the

layered model

investigate the relationship between electroseismic conversion

and

the conductivity of

the fluid saturated medium or the pore fluid. When fluid conductivity increases the

amplitude of

the

electroseismic wave increases. Experimental results show

that

electro-

seismic conversion is different from the piezoelectric effect of quartz grains in sandstone

and

is closely related to the relative motion between the fluid and

the

solid.

The

results

also eliminate

the

possibility

that

the electroseismic wave is generated by a spark of a

high voltage pulse. Electroseismic waves can be generated

at low

voltage

and

increased

continually with

the

voltage without a big voltage

jump

similar to

the

spark in an

isolated material.

Our

results confirm the existence

and

measurability of electroseismic

conversion

in

porous formation at ultrasonic frequency ranges. Therefore electroseis

12 1

7/26/2019 1997.12 Zhu et alnm,m

2/18

Zhu et al

mic measurements may be an -effective means to investigate the pore fluid

flow

and rock

properties.

INTRO U TION

In a porous medium most grain surfaces that are in contact with a fluid electrolyte

acquire a chemically-bound surface charge that is balanced by a diffuse distribution of

mobile counter ions in a thin fluid layer surrounding each grain. hebound-

and

diffuse

charged layers together are known as the electric double layer Broz and Epstein, 1976;

Cerda and Kiry, 1989; Chandler, 1981; Ishido and Mizutani, 1981; Kozak and Davis,

1989 .

the counter ions of the diffuse layer move by an electric force, the fluid will

move also. herelative movement and interaction between the fluid and the solid matrix

generate seismic vibration and mechanic waves Haartsen, 1995; Haartsen and Pride,

1996; Morgan

et

aI 1989

n this paper,

w

study the electroseismic waves generated by an electric pulse in

Berea sandstone cylinder models, confirming that electroseismic conversion exists in nat

ural, fluid-saturated porous rocks.

investigate the electroseismic waves generated by

the electrodes on the surface ofor inside the porous medium with a layered model made

of Lucite and epoxy_glued sand.

he

relationships between the

saturant

conductivity

and the electroseismic conversion, and the effects of the amplitude and polarization of

the

electric pulse on the electroseismic conversion, are studied experimentally.

EXPERIMENT IN A YLIN ER MODEL

nour experiments, a high-voltage 50-1000 V electric pulse, as an exciting source, is

applied to a fluid-saturated porous medium through electrodes with different shapes.

heround or square cylinder model is made of natural Berea sandstone. Ring or parallel

electrodes are attached on one side of the cylinder and connect to the high-voltage pulse.

A compressional P- or shear 3- transducer is fixed on the other side and receives the

seisn ic waves

Figure 1a is a diagram of the measurement system and the rock cylinder with ring

electrodes. When the high-voltage pulse connects with the nearest electrodes, the trans

ducer measures and records the compressional waves Figure 1b on the other side of the

cylinder. From the acoustic waveforms in Figure 1b,

w

see that the closer the exciting

area

is

to the transducer, the earlier the arrival time and the bigger the amplitude of the

wave. he frequencies of the waves are almost the same. This means that the electro

seismic waves are generated by the electric field along the axis in

the

area between

the

two electrodes. hiswave is an extensional wave propagating along the axis. Because

the length of the exciting area

is

the same during the course of the experiments, the

frequency of the electroseismic waves will be constant.

Figure 2a shows the Berea sandstone cylinder model with parallel electrodes. Two

electrodes of

the

conducting glue are on

the

sides of the square cylinder. A shear wave

12-2

7/26/2019 1997.12 Zhu et alnm,m

3/18

Electroseismic Conversion

transducer is fixed with shear wave coupling on the other side; its polarization can be

changed in the horizontal plane. When the electrodes are connected to a high-voltage

pulse generator,

an

alternate electric field is excited in the direction perpendicular to the

axis of

the

cylinder,

and

the shear components of the generated electroseismic waves

are received by the S-wave transducer at three polarizations Figure 2b . To check

the phases of the waveforms, the influences of the high-voltage pulse the first one

and-a-half circles remain in the plot and have the same arrival times, amplitudes and

phases. Not only the amplitude,

but

also the phase of the electroseismic waves vary

with

the

polarization of

the

shear transducer. This means that

the

electroseismic wave

is generated in the area between the electrodes and propagates as a flexural wave along

the axis of the cylinder. The particle motion

is

perpendicular to the axis

and

parallel

to the exciting electric field.

EXPERIMENT IN

A LAYERED MODEL

To study the electroseismic wave generated by the electrodes inside a porous medium,

we

made a layered model with a Lucite block and epoxy-glued fine sand where some

electrodes are buried. Some parameters of the glued sand and Lucite were listed in

Zim l 1997 . The reason for using the Lucite block is to delay the arrival time of

the electroseismic wave and to avoid the strong electronic influence of

the

high-voltage

pulse.

When the electrodes are on the surface of the glued sand and excited by an electric

pulse Figure 3a , and the P-wave transducer moves along t he the Lucite surface with

an increase of 2 em trace the transducer receives the electroseismic waves shown in

Figure 3b. We replace the electrodes with a P-wave source transducer and record the

acoustic field on

the

Lucite Figure 4a . Comparing the electroseismic waves Figure

3b with

the

acoustic waves Figure 4b , we see that they have the same arrival time

and similar amplitude variation. This means that the electroseismic wave is generated

in

the fluid-saturated porous medium near the electrodes and propagates

as

an acoustic

wave in the model.

the electrodes are at the same horizontal level Figure Sa ,

bu t

the spacing between

the negative and positive electrodes varies, the P-wave transducer generates electroseis

mic waves shown in Figure 5b when the spacings are 4 em, em, 8 em, 6 em, and 4

em, respectively. When the spacing decreases, the strength of the electric field between

the two electrodes increases and the generated electroseismic wave becomes stronger.

Although

the

distances between the center of

the

electrodes and

the

transducer are the

same,

the

arrival times of the electroseismic waves Figure 5b are different. This means

that

the depth or area where the electroseismic wave

is

generated increases or

is

closer

to the

transducer

when the electrodes are closer to each other and the s trength of the

electric field increases.

the spacing of the electrodes

is

fixed

at

different horizontal

levels Figure 6a , the electroseismic waves Figure 6b are almost the same. However,

the arrival time

and

the amplitude vary with the

depth

of the electrodes. From Fig-

12-3

7/26/2019 1997.12 Zhu et alnm,m

4/18

Zhu

et

al

ure 6b

we

calculate the velocity of the electroseismic wave in a fluid-saturated porous

medium glued sand , which is the P-wave velocity.

The above experiments with the layered model confirm that the electroseismic wave

is generated by

the

electric field between

the

electrodes and, as an acoustic wave, prop

agates in the medium.

PROPERTIES

ELE TROSEISMI

ONVERSION

The

above results show

that

the alternate electric field induces

the

relative motion be

tween the pore fluid and the solid matrix in a fluid-saturated porous medium. Their

interaction generates an acoustic wave known as an electroseismic wave.

This

phe

nomenon is essentially

the

conversion between electrical and mechanical energies.

To investigate the properties of the electroseismic conversion, we perform additional

experiments

with

a Berea sandstone layered model Figure 7 .

The

spacing between

the

two centers

ofthe

electrodes

is

8 em. The acoustic receiver is a P-wave transducer. With

this model

we

change

the

amplitude and the polarization

ofthe

exciting electric pulse, or

the conductivity of the saturant and record the electroseismic waves to investigate the

relationship between

the

electroseismic wave

and

the exciting pulse or

the

conductivity

of the pore fluid. Experiments also show the difference between the electroseismic

conversion

and

the piezoelectric effect or electric spark.

To study the relationship between the electroseismic conversion

and

the electric

conductivity of

the

porous medium,

we

cleaned the Berea sandstone block many times

with

pure

water in a vacuum system. We can then change the conductivity of the

water-saturated sandstone by changing that of the

saturant

water.

Our

experiments

recorded

the

electroseismic waves generated by

the

same electric pulse in the layered

model

saturated

by water with a different conductivity. Figure 8 shows

the

relationship

between the normalized amplitude of the electroseismic wave and the conductivity of the

saturant water . When the conductivity increases, the amplitude of

the

electroseismic

wave also increases. The opposite occurs in seismoelectric conversion Zhu

and

Toksiiz,

1996 , where the amplitude of the seismoelectric signal generated by a seismic wave

decreases when

the

conductivity of the saturant increases. This result also shows that

the electroseismic conversion is closely related to the pore fluid in the medium, because

the

interaction of

the

pore fluid

flow

on

the

solid matrix generates

the

electroseismic

w ve

The principle of electroseismic conversion is completely different

than

that of

the

piezoelectric effect. There are many quartz grains in sandstone. When an electric field

affects these grains,

it

generates an acoustic wave due to the piezoelectric effect of the

quartz. The

quartz

grains are very small

and

arranged randomly, thus the generated

acoustic wave is very weak or unmeasurable. We record the piezoelectric acoustic wave

generated by an 800 V electric pulse in a dry sandstone sample shown in Figure 7

and

compare it with the waves generated by the same pulse in the sandstone saturated with

water of 400

LS

in conductivity Figure 9 .

We

see

that

the acoustic wave due to the

12-4

7/26/2019 1997.12 Zhu et alnm,m

5/18

Electroseisrnic Conversion

piezoelectric effect

is

very weak compared with

that

of electroseismic conversion.

The

principle of electroseismic conversion

is

also different

than

that of an electrical

spark. When the voltage between two electrodes in an isolated medium increases to a

certain value, the medium may be electrically punctured,

and

a strong temporal current

passes through the medium and generates a spark and sound, resembling lightening and

thunder in the sky. The spark, caused by the sudden change of the electric property

of the medium, happens in an instance or does

not

vary continually with the voltage.

Figure

10

shows the relationship between the amplitude of the electric pulse and the

normalized maximum amplitude of the electroseismic wave when the pulse amplitude

varies

step

by step from

100

V to

1000

V in

the

model shown in Figure

In

the

voltage range of our experiments, the amplitude of the generated electroseismic

wave

varies continually with

that

of the electric pulse. is clear

that

the electroseismic

wave can also be received

at

a lower voltage range

100-200

V . This means that the

electroseismic conversion

is

not an electric spark. Electroseismic conversion does not

change

the

chemical properties of the medium and can only generate an ion movement in

the pore fluid. Electric signals with any amplitude can induce electroseismic conversion

and the difference is only in the amplitude of the generated electroseismic waves.

CONCLUSIONS

The

experimental results discussed here show

that

electric signals can generate electro

seismic waves with different modes in water-saturated cylinders or layered models. The

basic principle of electroseismic conversion is based on

the

motion of

the

ions in pore

fluid under the forces of an electric field. This movement induces fluid

flow

relative to

the solid matrix. The interaction between the fluid and the matrix generate seismic

waves that propagate in the medium.

Electroseismic conversion relates not only to pore fluid

flow

but

to

the

conductivity

of the fluid as well. The more ions there are in the pore fluid, the larger the ampli tude

of the electroseismic wave generated by the same electric pulse. This phenomenon

is

opposite to that of seismoelectric conversion, where more ions would counteract the

current of the

seismoelectric conversion.

The

results of

the

experiment show

that

the mechanism of the electroseismic con

version is completely different from

that

of the piezoelectric effect or electric spark.

Comparing the electroseismic conversion, the piezoelectric effect is very weak due to

the random

arrangement of

the

quartz grains in rocks. The continual variat ion of the

electroseismic conversion with the amplitude ofthe electric pulse shows

that it

is

not

an

electrical spark. The measurement of the electroseismic conversion would be an effective

method to explore pore fluid

flow

and some properties of porous formation.

2 5

7/26/2019 1997.12 Zhu et alnm,m

6/18

hu

et

al

KNOWLEDGMENTS

We

would like to thank Dr. A.

Thompson Prof. S Pride

and

Prof. T. Madden

for

their

valuable suggestions

and

useful discussions. This

study

is

supported

by

the

Borehole Acoustics and Logging Consortium at M.LT.

and

the

epartment

of Energy

rant No. DE FG02 93ER14322.

12 6

7/26/2019 1997.12 Zhu et alnm,m

7/18

Electroseismic

Conversion

REFERENCES

Broz

Z

an d N. Epstein 1976 Electrokinetic flow

through

porous

media

composed of

fine cylindrical capillaries

of

Colloid and Interface Science

56 605 612.

Cerda C.M. an d Non Chhom Kiry 1989 Th e use

of

sinusoidal

streaming

flow mea-

sur ements to

determine

th e

electrokinetic properties of porous media

Colloids and

Surfaces 35 7 15.

Chandler,

R ,

1981 Transient streaming potential measurements on fluid

saturated

porous structures: An experimental verification of Biot s slow wave in

th e

quasi-

static limit J Acoust. Soc. Am 70 116 121.

Haartsen,

M.W. 1995 Coupled electromagnetic an d acoustic wavefield modeling in

poro elastic media an d it s applications in geophysical exploration

Ph.D.

thesis MIT.

Haartsen, M.W. an d S.R. Pride 1996 Modeling of coupled electroseismic waves from

point sources in layered media submitted),

Geophys. Res.

Ishida

T. and

H. Mizutani 1981

Experimental and

theoretical basis

of

electrokinetic

phenomena

in rock water system an d its applications

to

geophysics

Geophys.

Res. 86 1763 1775.

Kozak M.W. an d

E.J.

Davis 1989 Electrokinetics of concentrated suspensions an d

porous media Colloid

of

and Interface Science 127 497 510 an d

of

Colloid

and Interface Science 129 166 174.

Morgan F.D. E .R Williams an d T .R Madden 1989 Streaming potential properties

of

westerly granite

with

applications

Geophys. Res. 94

12.449 12.461.

Zhu Z an d M.N. Toks6z 1996 Experimental studies of seismoelectric conversion in

fluid saturated porous medium

SEG 66th Annual International Meeting Expanded

Abstracts RP1.6 1699 1702 Denver CO.

Zhu Z. M.W. Haartsen an d M.N. Toks6z 1997 Experimental studies of electroseismic

conversions in a fluid saturated porous medium

submitted to Geophysics.

12-7

7/26/2019 1997.12 Zhu et alnm,m

8/18

Zhu et

al

Pulse

5

400V

1

em

C

0

-

J

l

1

iii

a.

(a)

E

m

lJ

0

II :

P-Wave

Receiver

5

c

4

0

c,

Q

x

3

Cl

c

.

2

u

x

w

1

b

o 50

100 150

200 250

Time

psec

Figure

Diagram for measuring the electroseismic waveform generated

by

an electric

pulse with ring electrodes a , and the electroseismic waves

b

received

by the

P-

wave transducer.

12-8

7/26/2019 1997.12 Zhu et alnm,m

9/18

Electroseismic

Conversion

E

o

o

I

I

105 CIJl

I

I

I

I

I

I

I

I

/

/

/

E

ctl

o

o

a:

Pulse

4 v

lectrode

ave

Receiver

a

-.. .----

3 y 1

2

Electric noise

5000000

00

wave

100

3

c:

0

.

0;;;

0

a.

2

b

Q)

>

u

Q

a::

1

Time

IlSec

Figure

2:

Diagram for measuring the electroseismic wave generated by an electric pulse

with parallel electrodes a), and

the

electroseismic waveforms b) received by the

S-wave transducer when its polarization

is at 1, 2, and 3

positions.

12-9

7/26/2019 1997.12 Zhu et alnm,m

10/18

Zhu et

al.

Electric pulse

500V

+

P-wave

Receiver

Receiver Position 2cm/trace

5 10

O --- ---- ---- --- ----\-- -.... ........ ....---. .-- --.

0.05

s0 10

E

:;::;

0.15

0.20

a

b

Figure

Glued-sand layered model a

and

the electroseismic waveforms b recorded

by fixing

the

electrodes on the top surface and moving the P-wave receiver with 2

em trace The

ampli tude is normalized by the maximum in

the

plot.

12-10

7/26/2019 1997.12 Zhu et alnm,m

11/18

Electroseismic Conversion

wave

source

a

e

wave

Receiver

Receiver Position 2cm/trace

5 10

b

>

-

r

?

I

0.05

o

0.20

0.15

J

0

.

10

Q

E

Figure Glued-sand layered model a and the acoustic waveforms b recorded by

fixing the source transducer on the top surface and moving the P-wave receiver with

2 cm/trace. The amplitude is normalized by the maximum in the plot.

12-11

7/26/2019 1997.12 Zhu et alnm,m

12/18

Zhu

e t

al.

Electric pulse

+

500 V)

Glued

sand

(a)

Lucite

.-/// ////,/,/,/1 /..- ,/ /

,

/// ///

,

/ 1 / / / / /

, , ,

////// / //// / //

, , ,

.................. , , , , ,

./ ./ ./ ./ . ./ . / .

.................. , , , ,

// / /

..............................................................................

/ / / /

./ . / ., , , ,

/

.I / / .; .I .I / ./

............................................................................

./

P-wave

Receiver

0.20.15

.1 0

time (ms)

0.05

c

0

4

E

l

0

0

Q)

b)

0

-

Q)

2

UJ

Figure

5:

Glued-sand layered model (a) with horizontal electrodes inside the glued sand

an d

th e

electroseismic waveforms (b) when the spacing of the electrodes varies from

em to 4 em with 4

em/trace.

12-12

7/26/2019 1997.12 Zhu et alnm,m

13/18

Electroseismic Conversion

lectric pulse

+

SOO V

Glued

sand

a

Lucite

P wave

Receiver

0.20

.15

.10

t ime ms

0.05

c

0

4

:E

f

0

b

0

-

2

w

Figure

Glued-sand

layered model a

with

horizontal electrodes inside the glued

sand

and the electroseismic waveforms b when the spacing of tIle electrodes varies with

1 em/trace.

12-13

7/26/2019 1997.12 Zhu et alnm,m

14/18

hu

et

at

lectric pulse

1

00 1

OOOV

5cm

5cm

P wave

Receiver

Figure

7:

Sandstone layered model

12 14

7/26/2019 1997.12 Zhu et alnm,m

15/18

1

Electroseismic

Conversion

-

ft:.

.................HHH HHHH

.

Q

C

0.8

E

0.6

C

Q

0.4

N

l

E

0.2

0

z

.

.H

............................., .

,.

. .

.

Water-saturated

sandstone

o

o

200

400 600

800

1000

aturant Conductivity J.IS

Figure

Relationship between the saturant water conductivity

and

the amplitude of

the electroseismic waveforms in the sandstone layered modeL

12-15

7/26/2019 1997.12 Zhu et alnm,m

16/18

hu al

1

2

o

0.05 1 15

Time ms

0.2

Figure

Waveforms recorded in the sandstone layered model when the sandstone is

ry 1

or s tur te

2

with

water of 400

f

in conductivity.

12-16

7/26/2019 1997.12 Zhu et alnm,m

17/18

Electroseismic Conversion

1

Ql

0

Co

E

0

Ql

N

0 8

0 6

0 4

ctl

E

o

z

O Z

o

o

zoo

6

olt ge

800

V

1000 1

Figure 1 : Relationship between the voltage of the exciting electric pulse

and

the am-

pli tude of th e electroseismic waveforms in

the

sandstone layered model

12 17

7/26/2019 1997.12 Zhu et alnm,m

18/18